FTIR spectroscopic semi-quantification of iron phases: A new method to evaluate the protection ability index (PAI) of archaeological artefacts corrosion systems

1 1FTIR spectroscopic semi-quantification of iron phases: a new 2method to evaluate the Protection Ability Index (PAI) of 3archaeological artefacts corrosion systems 4Marco Veneranda1*, Julene Aramendia1, Ludovic Bellot-Gurlet2, Philippe 5Colomban2, Kepa Castro1, Juan Manuel Madariaga1,3 61 Department of Analytical Chemistry, Faculty of Science and Technology, University of the 7Basque Country UPV/EHU, P.O. Box 644, 48080 Bilbao, Basque Country, Spain. 8[email protected] 92 Sorbonne Universités, UPMC Université Paris 6, MONARIS ‘de la Molécule aux Nano10 objets: Réactivité, Interactions et Spectroscopies’, UMR 8233, CNRS, 4 Place Jussieu, 75005, 11 Paris, France 12 3 Unesco Chair of Cultural Landscapes and Heritage, University of the Basque Country 13 (UPV/EHU), P.O. Box 450, 01080, Vitoria-Gasteiz, Spain 14 Abstract: 15 This study proposes an innovative approach to semi-quantify the main iron corrosion 16 phases found in corrosion systems of archaeological artefacts. This method is based on 17 the treatment of Fourier Transform Infrared Spectroscopy (FTIR) data using a 18 homemade spectra decomposition software (PALME). Its application was first tested on 19 mixtures of pure iron corrosion standards. After optimization, it was used to study real 20 archaeological samples and evaluate the stability of their corrosion system. 21 Considering that reliable and repetitive results were reached using extremely small 22 quantities of material, this method can be particularly suitable for the study of iron23 based objects of cultural interest. 24 Keywords: Iron corrosion; Semi-quantification; Fourier Transform Infrared 25 Spectroscopy; PALME software; Spectra decomposition; Archaeological artefacts; 26 1 Introduction: 27 The indispensable condition to ensure proper conservation of ancient iron artefacts is to 28 reach a state of chemical and physical balance with the environment in which they are 29 preserved [1]. This is the accepted manuscript of the article that appeared in final form in Corrosion Science 133 : 68-77 (2018), which has been published in final form at https://doi.org/10.1016/j.corsci.2018.01.016. © 2018 Elsevier under CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/) 2 1In this sense, the recovery of irons from archaeological sites only takes place when their 2matrix reaches the equilibrium with the soil, maintaining it throughout the burial time. 3However, this delicate balance suffers a critical disruption during archaeological 4excavation. In fact, artefacts recovery exposes them to a completely different 5environment, resulting often in the activation of many degradation pathways [2]. 6Therefore, conservators must act promptly in the later stages of archaeological 7excavation with the purpose of achieving, through specific conservation treatments, a 8renewed equilibrium with the new environmental context [3,4]. 9In this light, the work of conservators can be strongly benefited from analytical studies. 10 In the early stage, the molecular analysis of iron artefacts enables the identification of 11 the corrosion phases developed during the burial time. Those qualitative data are 12 extremely important to conservators, helping them to identify the real preservation state 13 of the analysed object. This is because each iron corrosion phase has a different 14 influence on the conservation of artefacts. With regards to archaeological iron artefacts 15 resumed from oxic environments, the corrosion system is generally mainly composed of 16 magnetite (Fe3O4), goethite (α-FeOOH), lepidocrocite (γ-FeOOH) and akaganeite (β17 FeOOH). In this context, is it well known that magnetite and goethite are stable 18 compounds that help the preservation of findings, whereas lepidocrocite and akaganeite 19 can be considered as degradation accelerators, as seen in the literature [5-7]. 20 In addition to qualitative analysis, further useful information can be obtained by 21 determining the relative concentration (quantitative or semi-quantitative analysis) of 22 each iron phase composing the corrosion system. For example, the quantitative analysis 23 of fragments that, one after the other, are removed by conservators during the 24 conservation work helps to identify the in-depth distribution of the reactive degradation 25 products and to characterize the reactivity of the whole corrosion system [8]. Such 26 approaches were also developed in the context of atmospheric corrosion in which the 27 quantification of iron phases was been to describe the stability of the corrosion layers 28 [9,10]. 29 The quantitative analysis of iron corrosion can also find reliable applications in the 30 stages following the conservation works. In the short term, it can be used to control 31 whether conservation treatments were able to stabilize the reactive corrosion 3 1phases[11]. In the long term, it can also be used to identify any new corrosion processes 2that may be entailed by the interaction between the object and the storage/exhibition 3environmental conditions [12,13]. 4In this context, molecular analytical techniques such as X-Ray Diffraction (XRD) 5[14,15] and Mossbauer spectroscopy [16,17] have been extensively used with the 6specific purpose of determining the relative concentration of the phases that compose 7the iron corrosion systems. 8Morevover, Raman spectroscopy is acquiring a steadily increasing importance in studies 9related to cultural heritage materials due to its versatility and capability of collecting 10 molecular data in a non-intrusive way[18]. 11 Considering that the intensity of Raman signals are proportional to the concentration, 12 the most used quantification method is the one based on the use of external calibration 13 curves [19-21]. To avoid the use of calibration curves for each analysed compound and 14 their mixing, an approach using spectral decomposition in a linear combination of 15 reference spectra was proposed for studying atmospheric corrosion of medieval iron 16 [10]. To go further than using point analyses the corrosion heterogeneities were taken 17 into account using the automated treatment of Raman maps over the corrosion system. 18 This approach was also applied for the diagnostic of self-weathering steel atmospheric 19 corrosion of contemporary work of Art [22]. 20 However, it must be pointed out that Raman spectroscopy is poorly suitable for the 21 study of materials featuring high auto-fluorescence emissions. In these cases, the use of 22 Fourier Transform Infrared Spectroscopy (FTIR) is more indicated, since it avoids any 23 problem related to the auto-fluorescence of the sample. Moreover, working with 24 powdered samples, FTIR systems also ensure a better accuracy and repeatability of the 25 results over the sample heterogeneities by sampling the whole corrosion system [23]. 26 Even though FTIR systems have been successfully applied for the quantification of 27 several kind of liquid [24,25] and solid [26,27] samples, only a few works describe the 28 use of this spectroscopic technique for the quantification of iron phases [28, 29]. 29 Considering that this analytical approach has never been applied in the field of cultural 30 heritage characterization, the main objective of the present work was to evaluate if FTIR 4 1spectroscopy can be used as an alternative technique to semi-quantify the main 2corrosion phases of iron archaeological artefacts. For this purpose, a dedicated software 3(PALME) designed by the LADIR group (now MONARIS, Pierre and Marie Curie 4University, France) was employed to perform the decomposition of FTIR spectra. 5Afterwards, the proposed procedure was also used to assess whether it can be used to 6reliably assess the stability of real rust samples coming from archaeological artefacts. In 7this regards, the stability assessment has been inspired by the protection ability index 8(PAI index), proposed for the first time by Yamashita et al. [30] and subsequently 9adjusted by Dillmann et al. [7] for the analysis of rust layers covering ancient iron 10 objects exposed to atmospheric corrosion. 11 The main advantage obtained from PAI index calculation consists in helping 12 conservators on predicting the corrosion behaviour of iron artefacts after their recovery 13 from the archaeological site. Thus, objects providing high stability values can be treated 14 and stored by following routine protocols, whereas unstable artefacts require targeted 15 treatments in order to prevent the onset of post-excavation degradation processes. In this 16 context, it is important to clarify that the semi-quantification of iron phases involves the 17 sampling / processing of corrosion material. However, this aspect does not represent a 18 remarkable issue since thick corrosion systems are generally removed / thinned in order 19 to recover the original shape of the artefacts [31]. 20 2 Experimental methods and tests 21 2.1 Samples preparation 22 This work was based on the analysis of both, standard mixtures and archaeological rust 23 samples. In the first case, pure magnetite, lepidocrocite and goethite iron phases (from 24 Sigma-Aldrich corp. St Luis, USA) were kindly provided by D. Neff from the LAPA 25 group (NIMBE UMR3685 CEA/CNRS, France). On the other hand, pure akaganeite 26 was synthesized using the method described by Reguer et al. [32]. The synthesis method 27 involved the hydrolysis of a 0.1 M ferric chloride solution (FeCl3 • 6H2O) by heating 2 28 liters of the solution at 70 °C during 48 h. 29 On the one hand, 10 standard mixtures were prepared by mixing iron oxide and 30 oxyhydroxide standards at different proportions. Considering that FTIR spectroscopy 5 1needs a small particle size (1-2 μm) to avoid any distortion phenomena, an agate mortar 2was used for the grinding and the homogenization of all samples. The relative weight of 3each iron phase in the mixtures was monitored by using an analytical balance (AE200, 4Mettler) with an accuracy of 0.0001 g. 5On the other hand, real samples were collected from iron artefacts excavated in the 6Roman archaeological site of Forua (Spain) [33]. This Roman settlement, discovered in 71982, stands just few kilometres inland from the Bay of Biscay. In the archaeological 8excavations several objects were discovered, including five iron-based nails dated back 9between the 2nd and the 4th century A.D (see Figure SM1 in supplementary material). 10 Since their recovery, all artefacts have been constantly kept in a controlled environment 11 room (temperature and relative humidity of 20 ± 2 °C and 65 ± 2% respectively) 12 without the implementation of further conservation treatments. To minimize post13 excavation corrosion phenomena, each nail had been stored in hermetic boxes equipped 14 with desiccant silica gel beads that ensure humidity levels below 10% relative humidity. 15 In the frame of this study, one rust micro-sample was collected from each of the five 16 nails. The sampling, carried out with the collaboration of the conservators of the 17 Archaeological Museum of Bizkaia, was performed after cleaning the outer rust layer 18 from impurities (e.g. earth, clays and organic material) deposited on its surface during 19 the burial time. In this way, the possible interferences proportioned by extraneous 20 materials during the analytical characterization of the iron corrosion were minimized. 21 The collected samples were finally analysed to semi-quantify the main iron phases and 22 evaluate their stability through the protective ability index calculation. 23 2.2 FTIR systems 24 For the development of this work two FTIR systems were used. On the one hand, a 25 Jasco 6300 system, operating in transmittance, diffuse reflection (DRIFT, Jasco DR 26 PR0410M) and ATR (diamond crystal with a ZnSe focusing lens, PIKE Miracle™) 27 modes, was used with the aim of checking what configuration provided the most 28 reliable results. The instrument is equipped with a Ge on KBr beamsplitter, a Michelson 29 interferometer and a DLaTGS detector with Peltier temperature control. Analysis were 30 performed in the middle infrared region (from 4000 to 400 cm-1) recording 64 scans at 4 31 cm-1 spectral resolution. 6 1To collect ATR spectra, a small portion of homogenized sample (around 0.05 g) was 2placed in the microsample holder, firmly clamped against the ATR crystal and analysed 3in its pure form. 4On the other side, KBr-matrix pellets were made to carry out transmittance analysis. To 5prepare the pellets, 0.5 mg of sample was mixed with 170 mg of dry KBr (>99% FTIR 6grade, Sigma-Aldrich), milled in an agate mortar and pressed under 10 tons (CrushIR, 7PIKE technologies) for 8 minutes. 8For DRIFT analysis, the microsample holder was filled with a powder mixture 9composed of 10% (w/w) sample and 90% (w/w) KBr. This dilution ratio ensured having 10 a lower specular component on the surface of the sample increasing the contribution of 11 the diffuse reflectance component [23]. 12 In a second step, the results obtained by the use of the above described laboratory 13 instruments were compared with those of a portable FTIR. The aim was to verify if the 14 proposed semi-quantification method could be applied to in-situ analysis. For this 15 purpose, a compact portable Alpha FTIR spectrometer (Bruker Optics Inc., Germany) 16 equipped with a Ge on KBr beamsplitter and a diamond ATR accessory was used. 17 To compare the results of the portable instrument with those obtained by the laboratory 18 one, the same measurement parameters were used (64 acquisitions at 4 cm−1 resolution 19 over a spectral range of 400–4000 cm−1, see above). 20 To improve the reliability of the proposed method, the FTIR spectra obtained from both 21 portable and laboratory systems were treated using the Opus 7.2 software (Bruker 22 Optics, Germany). Thus, CO2/H2O and noise corrections, spectra baseline adjustment 23 and fingerprint region selection were performed. 24 2.3 PALME software 25 After completing the analysis of all samples by using both FTIR systems, the PALME 26 software (Program d'AnaLyse vibrationnelle de spectres de MElanges à partir de 27 spectres purs) developed by the LADIR Laboratory (now MONARIS, Pierre et Marie 28 Curie University, France) was applied to semi-quantify the detected iron corrosion 29 phases. This program was specifically designed to treat spectra provided by vibrational 30 spectroscopy techniques. PALME software automatically performs the semi- 7 1quantification of compounds mixtures by the linear combination of spectra of pure 2reference standards [34,35]. The process consists of two steps. In the first one, the 3software uses a sum of Gaussian and/or Lorentzian band profiles and a least-square 4fitting to produce for each reference compound (in this work akaganeite, lepidocrocite, 5goethite and magnetite) a calculated spectrum that fits the recorded experimental one 6[36]. The calculations and the fitting procedures performed by PALME software have 7been detailed in a specific publication [37]. 8In a second step, a linear combination of the calculated standard spectra is used for the 9fitting of a spectrum similar to the sample spectrum by means of the least-squares 10 criterion and the Levenberg-Marquardt algorithm. After validation of the fitting by the 11 user, PALME software provides a txt document including the contribution of each 12 standard (expressed as a weighting coefficient) to the decomposition of the sample 13 spectrum. 14 2.4 Corrosion system stability evaluation 15 After completing the semi-quantification of the iron phases, the percentage values of 16 each compound were used to determine the sample corrosion stability. 17 As explained above, the adjusted index (*PAI index) considers the presence of 18 akaganeite, lepidocrocite as reactive phases; and goethite, and magnetite as protective 19 ones (Equation 1). Unlike Yamashita et al. [30], due to its passivity and stability, 20 magnetite (Fe3O4) is here considered as protective. 21 In this light, the stability of archaeological corrosion samples was calculated as follows: 22 (Eq.1) Corrosion stability = mass fraction ( α FeO(OH) ) + (Fe 3 O 4 ) mass fraction ( γ FeO(OH) ) + ( β FeO(OH)) 23 In the case of the mixtures of standards, the reliability of the stability calculation was 24 determined by comparison with the results obtained using the weighted proportions of 25 each compounds. In the case of archaeological samples, the reliability evaluation was 26 performed using as a reference the value calculated through the semi-quantification 27 values obtained by X-ray diffraction (XRD) analysis. 28 2.5 X-ray diffraction 8 1In order to characterize and semi-quantify the iron phases in the archaeological rust 2samples, a PRO PANalytical Xpert XRD was used. 3The system is equipped with a copper tube, a vertical goniometer (Bragg-Brentano 4geometry), a programmable divergence slit, a secondary graphite monochromator and a 5Pixcel detector. The condition of all measurements were set at 40 KV, 40 mA and a 6scan ranging between 5 and 70º 2theta. 7It is important to recall that the peaks’ shape of the XRD diffractogram depends on the 8crystallinity of the phases (the narrowness of the peaks increase with increasing short 9and long range ordering). For this reason semi-quantitative values were obtained by 10 treating areas values. 11 To obtain semi-quantitative data from the collected diffractograms, two different 12 softwares were used. Both Xpert HighScore and EVA software (PANanalytical, 13 Holland) apply the Reference Intensity Ratio (RIR) method to predict the phase 14 abundances. Considering that the intensity of a diffraction peak profile is a convolution 15 of many factors, the RIR method measures and reduces to a constant all the factors 16 except concentration to determine phases concentration (by comparison to a reference 17 pattern). Although the softwares are both based on the Reference Intensity Ratio 18 method, the algorithm used for the decomposition of the experimental data is different, 19 which can result in different semi-quantitative results. 20 3 Results and Discussion 21 3.1 Characterization of pure standards 22 The semi-quantification performed by PALME software is based on the decomposition 23 of the vibrational spectrum of a sample by comparison with pure reference spectra. 24 Thus, the first step was focused on the collection of the characteristic FTIR spectra of 25 pure akaganeite, lepidocrocite, goethite and magnetite using the same experimental 26 conditions in order to obtain absorbances (relative intensities between bands) related to 27 each sample specificities, spectrometer characteristics and measurement mode 28 (transmittance, DRIFT, ATR). For example, Figure 1 reports the FTIR spectra of pure 29 iron phases, collected by using the JASCO 6300 laboratory system in transmittance 30 mode. 9 1 2Figure 1. FTIR raw spectra of iron oxide standards carried out using the Jasco 6300 3system in transmittance mode. 4As showed in Figure 1, the magnetite vibrational spectrum was characterized by the 5presence of strong signals at 588 and 3437 cm-1 (Figure 1a). Goethite spectrum showed 6three strong peaks at 615, 798 and 905 cm-1 respectively, together with two broad bands 7at 3136 and 3431 cm-1(Figure 1b). Lepidocrocite spectrum stood out by the presence of 8a main peak at 1023 cm-1, followed by several secondary signals at 485, 615, 759, 1152, 93014 and 3414 cm-1(Figure 1c). Finally, akaganeite standard provided an intense double 10 peak at 648 and 693 cm-1, together with two weak signals at 844 and 1623 cm-1 and a 11 broad band at 3385 cm-1 (Figure 1d). 12 To illustrate the signal differences produced by each acquisition mode, the comparison 13 of the akaganeite spectra recorded using the Jasco 6300 in ATR, DRIFT and 14 transmittance modes are shown in Figure 2. 15 16 Figure 2. Comparison among akaganeite spectra obtained using the Jasco 6300 system 17 in ATR, DRIFT and transmittance modes. 18 As evidenced in Figure 2, transmittance and DRIFT modes were more sensitive to the 19 O-H bonds vibrations with respect to the ATR mode [38,39], promoting the 20 enhancement of the hydroxide signals on the spectrum of pure akaganeite (3385 and 21 3470 cm-1). This is due to the lesser penetration of the high wavenumber IR lights in 22 ATR mode which result in a smaller volume sampled and then a lower absorbance. As 23 here the ATR spectra were not corrected by the ATR correction function which is 24 commonly provided by the IR software (we choose to not introduce such corrections of 25 spectra) the differences between the ATR profile and the transmittance/DRIFT ones 26 remain. Furthermore, it must be pointed out that, using the DRIFT method, the infrared 27 radiation penetrates the sample/KBr mixture, producing a high number of refractions 28 and absorption processes which result in an enhancement of the intensity in the 29 interfaces contribution, and combination bands [40]. This statement was confirmed by 30 experimental data owing to the presence of several secondary peaks in a wavelength 16 1Sample 3) require specific stabilization treatments (such as desalination baths) that 2minimize post-excavation degradation issues. 34 Conclusions 4This paper introduces a new analytical method, based on the use of FTIR spectroscopy, 5to evaluate the stability of archaeological artefacts corrosion systems. 6In the first step, the semi-quantification of standard mixtures spectra collected by means 7of laboratory system (Jasco 6300) indicated that the most reliable results were provided 8as expected by the decomposition of transmittance spectra (powder dispersed in a KBr 9pellet). This work also demonstrated that, by performing spectra treatments (fingerprint 10 region selection and baseline correction), the method accuracy especially of those using 11 DRIFT or ATR procedures can be significantly improved. 12 Afterwards, the results obtained by means of laboratory systems (Transmission, DRIFT 13 & ATR) were compared to those of the portable Alpha IR spectrometer in ATR mode. 14 As proved by the trend line showed in Figure 4, the results obtained by semi15 quantifying treated Alpha spectra were good enough to enable the application of the 16 proposed protocol also to on site analysis. Once the analytical procedure was validated, 17 the current availability of various small movable ATR-FTIR devices allows working out 18 of analytical laboratories, in restoration workshops, museums or even on the 19 archaeological field. 20 PALME concentration values obtained by the decomposition of both portable (in ATR 21 mode) and laboratory (in transmittance mode) spectra were used to calculate the 22 stability values of both standard mixtures and archaeological samples. Concretely, the 23 experimental stability values obtained from archaeological iron corrosion samples were 24 in line with those calculated from XRD data. For this reason we believe that the FTIR25 PALME method represents a viable alternative to those based on the treatment of X26 Ray diffractograms for both iron phases semi-quantification and stability evaluation. 27 Finally, it must be emphasized that FTIR analyses require a smaller amount of sample 28 (0.3 - 0.5 mg in transmittance mode) compared to the XRD technique. Thus, this 29 method can be particularly suitable for the study of iron-based objects of cultural 17 1interest because the amount of sample that is usually available is very little due to the 2special characteristic of the cultural objects. 3In conclusion, the reliability of the proposed method for PAI index determinations could 4represent a great advantage for conservators, helping them on predicting the corrosion 5behaviour of iron artefacts and consequently planning optimal conservation treatment. 6Acknowledgments 7This project has been funded by the UFI "Global Change and Heritage" project (Ref 8UFI 11-26 UPV-EHU), the DISILICA-1930 project (ref BIA2014-59124) and the 9European Regional Development Fund (FEDER). Marco Veneranda thanks the 10 Ministry of Innovation and Competitiveness (MINECO) for his pre-doctoral fellowship. 11 We would like to thank the LAPA laboratory for providing iron phases standards and 12 Professor Juan Manuel Gutierrez Zorrilla for his support in the synthesis of akaganeite. 13 References 14 [1] M. Veneranda, J. Aramendia, O. Gomez, S. Fdez-Ortiz de Vallejuelo, L. Garcia, I. 15 Garcia-Camino, K. Castro, A. Azkarate, J.M. Madariaga, Characterization of 16 archaeometallurgical artefacts by means of portable Raman systems: corrosion 17 mechanisms influenced by marine aerosol, J. Raman Spectrosc. 48 (2016) 258-266. 18 [2] S.Turgoose, Post-excavation changes in iron antiquities, Stud. Conserv. 27 (1982) 19 97-101. 20 [3] D. Ashkenazi, I. Nusbaum, Y. Shacham-Diamand, D. Cvikel, Y. Kaanov, A. Inberg, 21 A method of conserving ancient iron artefacts retrieved from shipwrecks using a 22 combination of silane self-assembled monolayers and wax coating, Corros. Sci. 123 23 (2017) 88-102. 24 [4] Ph. Dillmann, G. Beranger, P. Piccardo, H. Matthiessen, Corrosion of Metallic 25 Heritage Artefacts, Volume 48, first ed., Woodhead Publishing, England, 2007. 26 [5] M. Veneranda, I. Costantini, S. Fdez-Ortiz de Vallejuelo, L. Garcia, I. García, K. 27 Castro, A. Azkarate, J. M. Madariaga, Study of corrosion in archaeological gilded irons 18 1by Raman imaging and a coupled scanning electron microscope–Raman system, Philos. 2T. Roy. Soc. A. 374 (2016) 20160046. 3[6] J. Monnier, P. Dillmann, L. Legrand, I. Guillot, Corrosion of iron from heritage 4buildings: proposal for degradation indexes based on rust layer composition and 5electrochemical reactivity, Corros. Eng. Sci. Techn. 45 (2013) 375-380. 6[7] Ph. Dillmann, F. Mazaudier, S. Hoerlé, Advances in understanding atmospheric 7corrosion of iron. Rust characterization of ancient ferrous artefacts exposed to indoor 8atmospheric corrosion, Corros. Sci. 46 (2004) 1401-1429. 9[8] D. Neff, S. Reguer, L. Bellot-Gurlet, Ph. Dillmann, R. Bertholon, Structural 10 characterization of corrosion products on archaeological iron: an integrated analytical 11 approach to establish corrosion forms. J. Raman Spectrosc. 35 (2004) 739-745. 12 [9] M. Yamashita, T. Misawa, Recent progress in the study of protective rust-layer 13 formation on weathering steel, in: Proceeding of the NACE Corrosion Conference, 14 1998, Paper No. 357. 15 [10] J. Monnier, L. Bellot-Gurlet, D. Baron, D. Neff, I. Guillot, Ph. Dillmann, A 16 methodology for Raman structural quantification imaging and its application to iron 17 indoor atmospheric corrosion products, J. Raman Spectrosc. 42 (2011) 773-781. 18 [11] M. Bayle, P. de Vivés, J.B. Memet, E. Foy, Ph. Dillmann, D. Neff, Corrosion 19 product transformations in alkaline baths under pressure and high temperatura: The sub20 critical stabilisation of marine iron artefacts stored under atmospheric conditions, Mater. 21 Corros. 57 (2016) 190-199. 22 [12] S. Réguer, F. Mirambet, C. Rémazeilles, D. Vantelon, F. Kergourlay, D. Neff, Ph. 23 Dillmann, Iron corrosion in archaeological context: Structural refinement of the ferrous 24 hydroxychloride β-Fe2(OH)3Cl, Corros. Sci. 100 (2015) 589-598. 25 [13] M. Bouchar, Ph. Dillmann, D. Neff, New Insights in the Long-Term Atmospheric 26 Corrosion Mecanisms of Low Alloy Steel Reinforcements of Cultural Heritage 27 Buildigs, Meterials 10 (2017) 1-16. 19 1[14] M. Morcillo, B. Chico, J. Alcántara, I. Díaz, D. de la Fuente, SEM/Micro-Raman 2Characterization of the Morphologies of Marine Atmospheric COrrosion Products 3Formed on Mild Steel, J. Electrochem. Soc, 163 (2016) C426-C439. 4[15] D. de la Fuente, J. Alcántara, B. Chico, I. Díaz, J.A. Jiménez, M Morcillo, 5Characterisation of rust surfaces formed on mild steel exposed to marine atmospheres 6using XRD and SEM/Micro-Raman techniques, Corros. Sci, 110 (2016) 253-264. 7[16] D.C. Cook, Spectroscopic identification of protective and non-protective corrosion 8coatings on steel structures in marine environment, Corr. Sci 47 (2005) 2550-2570. 9[17] R.E. Vandenberghe, C.A. Barrero, G.M. da Costa, E. Van San, E. De Grave, 10 Mössbauer characterization of iron oxides and (oxy)hydroxides: the present state of the 11 art, Hyperfine Interact., 126 (2000) 247-259. 12 [18] N. Yucel, A. Kalkanli, E.N. Caner-Saltik, Investigation of atmospheric corrosion 13 layers on historic iron nails by micro-Raman spectroscopy, J. Raman Spectrosc. 47 14 (2016) 1486-1493. 15 [19] M. Veneranda, M. Irazola, M. Díez, A. Iturregui, J. Aramendia, K. Castro, J. M. 16 Madariaga, Raman spectroscopic study of the degradation of a middle age mural 17 painting: the role of agricultural activities, J. Raman Spectrosc. 45 (2014) 1110-1118. 18 [20] J.H. Giles, D.A. Gilmore, M.B. Denton, Quantitative analysis using Raman 19 spectroscopy without spectral standardization, J. Raman Spectrosc. 30 (1999) 767-771. 20 [21] F. Dubois, C. Mendibide, T. Pagnier, F. Perrard, C. Duret, Raman mapping of 21 corrosion products formed onto spring steels during salt spray experiments. A 22 correlation between the scale composition and the corrosion resistance, Corros. Sci. 50 23 (2008) 3401-3409. 24 [22] J. Aramendia, L. Gomez-Nubla, L. Bellot-Gurlet, K. Castro, C. Paris, Ph. 25 Colomban, J.M. Madariaga, Protective ability index measurement through Raman 26 quantification imaging to diagnose the conservation state of weathering steel structures, 27 J. Raman Spectrosc. 45 (2014) 1076–1084. 20 1[23] I. Arrizabalaga, O. Gómez-Laserna, J.A. Carrero, A. Rodriguez, G. Arana, J.M. 2Madariaga, Diffuse reflectance FTIR database for the interpretation of the spectra 3obtained with a handheld device on built heritage materials, Anal. Methods. 7 (2015) 41061-1070. 5[24] I. F. Duarte, A. Barros, I. Delgadillo, C. Almeida, A.M. Gil. Application of FTIR 6Spectroscopy for the Quantification of Sugars in Mango Juice as a Function of 7Ripening, J. Agric. Food Chem, 50 (2002) 3104–3111. 8[25] O. Anjos, M.G. Campos, P.C. Ruiz, P. Antunes, Application of FTIR-ATR 9spectroscopy to the quantification of sugar in honey, Food Chem. 169 (2015) 218-223. 10 [26] M.A. Legodi, D. De Waal, J.H. Potgieter, Quantitative determination of CaCO3 in 11 cement blends by FT-IR, Appl. Spectrosc. 55 (2001) 361–365. 12 [27] F.B. Reig, J.V.G. Adelantado, M.C.M. Moya Moreno, FTIR quantitative analysis 13 of calcium carbonate (calcite) and silica (quartz) mixtures using the constant ratio 14 method. Application to geological samples, Talanta. 58 (2002) 811–821. 15 [28] H. Namduri, S. Nasrazadani, Quantitative analysis of iron oxides using Fourier 16 transform infrared spectrophotometry, Corros. Sci. 50 (2008) 2493–2497. 17 [29] A.F. Betancur, F.R. Pérez, M.M. Correa, C.A. Barrero, Quantitative aproach in 18 iron oxides and oxihydroxides by vibrational analysis, Opt. Pura Apl. 45 (2012) 26919 275. 20 [30] M. Yamashita, H. Miyuki, Y. Matsuda, H. Nagano, T. Misawa, The long term 21 growth of the protective rust layer formed on weathering steel by atmospheric corrosion 22 during a quarter of a century, Corros. Sci. 36 (1994) 283-299. 23 [31] D. Ashkenazi, I. Nusbaum, Y. Shacham-Diamand, D. Cvikel, Y. Kahanov, A. 24 Inberg, A method of conserving ancient iron artefacts retrieved from shipwrecks using a 25 combination of silane self-assembled monolayers and wax coating, Corros. Sci. 123 26 (2017) 88-102. 27 [32] S. Reguer, F. Mirambet, E. Dooryhee, J.L. Hodeau, Ph. Dillmann, P. Lagarde, 28 Structural evidence for the desalination of akaganeite in the preservation of iron 21 1archaeological objects, using synchrotron X-Ray powder diffraction and absorption 2spectroscopy, Corros. Sci. 51 (2009) 2795-2802. 3[33] C. Fernández Ibáñez, Theatre bronce mask as ornamental top of a chandelier, 4coming from roman deposit in Forua (Biskay area), KOBIE. 6 (2004) 455-470. 5[34] F. Salpin, F. Trivier, S. Lecomte, C. Coupry, A new quantitative method: non6destructive study by Raman spectroscopy of dyes fixed on wool fibres, J. Raman 7Spectrosc. 37 (2006) 1403-1410. 8[35] C. Daher, V. Pimenta, L. Bellot-Gurlet, Towards a non-invasive quantitative 9analysis of the organic components in museum objects varnishes by vibrational 10 spectroscopies: methodological approach, Talanta.129 (2014) 336-345. 11 [36] I. Costantini, M. Veneranda, N. Prieto-Taboada, L. Bellot-Gurlet, K. Castro, J.M. 12 Madariaga, Comparison of semicuantification experimental methodologies using micro13 Raman spectroscopy: PALME software as an altertative tool for the study of salt 14 efflorescence, J. Raman Spectrosc. 47 (2016) 1415-1421. 15 [37] S. Döpner, P. Hildebrandt, A. Grant Mauk, H. Lenk, Werner Stempfle, Analysis of 16 vibrational spectra of multicomponent systems. Application to pH-dependent resonance 17 Raman spectra of ferricytochrome c, Spectrochim. Acta A, 51 (1996) 573-584. 18 [38] I. Arrizabalaga, O. Gomez-Laserna, J. Aramendia, G. Arana, J.M. Madariaga, 19 Determination of the pigments present in a wallpaper of the middle nineteenth century: 20 The combination of mid-diffuse reflectance and far infrared spectroscopies, 21 Spectrochim. Acta A. 124 (2014) 308-314. 22 [39] I. Arrizabalaga, O. Gomez-Laserna, J. Aramendia, G. Arana, J.M. Madariaga, 23 Applicability of a Diffuse Reflectance Infrared Fourier Transform handheld 24 spectrometer to perform in situ analyses on Cultural Heritage materials, Spectrochim. 25 Acta A. 129 (2014) 259-267. 26 [40] C. Miliani, F. Rosi, B. G. Brunetti, A. Sgamelloti, In situ noninvasive study of 27 artworks: The MOLAB multitechnique approach, Acc.Chem. Res. 43 (2010) 728–738. 22 1[41] Ph. Colomban, Hydrogen bonding in hydrogen-substituted lithium 2aluminosilicates, J. Mol. Struct. 270 (1992) 407-416. 3[42] T.H. Anderson, F.W. Weaver, N.L. Owen, Anomalies in diffuse reflectance 4infrared spectra of wood and wood polymers, J. Mol. Struct. 249 (1991) 257-275. 5[43] R. Tisserand, M. Rebetez, M. Grivet, S. Bouffard, A. Benyagoup, F. Levesque, J. 6Carpena, Comparative amorphization quantification of two apatitic materials irradiated 7with heavy ions using XRD and RSB results, Nuc. Instrum. Meth. A. 215 (2004) 1298136. 9[44] P. Blanc, O. Legendre, E.C. Gaucher, Estimate of clay minerals amounts from 10 XRD pattern modeling: The Arquant model, Phys. Chem. Earth. 32 (2007) 135-144. Highlights:  A new iron corrosion semi-quantification method is proposed.  A homemade decomposition software (PALME) was used for the treatment of Fourier Transform Infrared Spectroscopy (FTIR) spectra.  This approach was applied for the study of both, mixtures of pure iron corrosion standards and real archaeological corrosion samples  Fast, reliable and repetitive results were reached using extremely small quantity of material (below 0.5g).